PROJECT SUMMARY Glioblastoma (GBM) is a lethal, incurable form of cancer in the brain. Even with maximally aggressive surgery and chemoradiotherapy, median survival is 14.5 months. These tumors infiltrate normal brain tissue, are surgically incurable, and universally recur. At recurrence, tumors are often resistant to standard treatments with response rates of less than 5% and a median survival of 8 months. GBMs are characterized by genetic, epigenetic, and microenvironmental heterogeneity, and they evolve spontaneously over time and as a result of treatment. During these evolutions, the heterogeneous makeup can shift, both phenotypically and genetically, which can include changes in drug responsiveness, making selection of therapies at any given point in time a challenge. In this proposal, we seek to address these challenges by biofabricating ex vivo 3D organoid models of GBM for specific patients that will recapitulate intratumoral heterogeneity and track clonal evolution in response to patient-specific therapy. Towards this goal, we have developed a platform of biofabricated tumor organoids, including a GBM-on-a-Chip system. Our studies have shown that GBM subpopulations of different genetic profiles respond to external stimuli differently, and shift in terms of organoid percent makeup. Additionally, we have developed a robust system for creating 3D tumor organoids from patient biospecimens with over a 80% take rate ? compared to 25-33% take rates in 2D cultures and patient-derived xenograft models ? in a variety of cancer types, including colon, appendiceal, mesothelioma, sarcoma, ranging from low to high grade malignancies. With this library of patient organoid sets we have demonstrated capabilities to perform drug screening studies, including verifying correlation with patient response as well as biomarker- driven drug tests. This R33 proposal seeks to employ this technology to create 3D ex vivo GBM organoids on- a-chip from GBM patient tumor biospecimens, and will probe subtype population drift and genetic profile changes over time in parallel to the patient, following the same therapeutic interventions administered to the patient. Aim 1 will deploy these patient-specific organoids and assess subtype population distribution and evolution, genetic profiles through sequencing, and phenotype over time. Aim 2 will use the same output metrics, but will administer the patient-specific treatment (together with no treatment controls), thereby assessing how the tumor organoids evolve after clinically relevant therapeutic intervention. Significant outcomes of this work, proposed by an early-stage investigator-led team, will be a versatile ex vivo technology that supports patient tumor ?avatars? for tracking in parallel how a particular patient's tumor is changing over time, providing a tool that mimics patient tumor evolution, which can be deployed to predict which therapies are most effective for patients at a given point in time for this currently incurable disease.